The Next Frontier of Precision Medicine: Spatial Biology Meets Genomics
Cellular Biomechanics in Drug Resistance Detection and Spatial Biomedical Analysis
Cellular traction force measurements using high-resolution microscopy to quantify pillar displacements have been widely applied to study how forces are influenced by drugs, mutations, or gene knockdowns. However, conventional micro- and nanopillar arrays, which require high magnification imaging, are limited to single-cell studies and not suitable for high-throughput screening. Here, we present the Force Arrays Screening Tool (FAST), a label-free, high-throughput traction force imaging platform that measures reflected light intensity from metal-coated pillars rather than displacements. FAST enables rapid mapping of traction force distributions across thousands of cells within seconds at single-cell resolution. We show that traction force signatures can serve as sensitive biophysical biomarkers to assess drug cytotoxicity, distinguish cancer from normal cells, and identify drug-resistant subpopulations. Importantly, traction force changes correlated with YAP and β1-integrin activity, key mediators of mechanotransduction. In drug-resistant lung cancer cells, these molecules sustained resistance to epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors. Inhibition of YAP or β1-integrin restored both mechanical phenotypes and drug sensitivity. FAST allows mapping of over 100,000 cells per minute, bridging the gap between single-cell mechanobiology and scalable screening. This technology reveals novel biomechanical signatures of disease states with potential applications in diagnostics and therapeutic discovery.
Structure-guided machine learning prediction of the relationship between single amino acid variations, cancer, and drug resistance
Single amino acid variation (SAV) refers to a substitution of an amino acid in a protein sequence, which can potentially affect the overall structure and function of the protein, as well as its binding affinity. Protein destabilization is linked to various diseases, including several types of cancer, and SAVs may contribute to resistance against anticancer drug therapy. In our study, we transformed all characteristics of SAVs derived from protein sequences, structures, and their microenvironments into feature vectors. These vectors were then processed through an integrated prediction system utilizing a support vector machine and genetic algorithms. We focused on identifying critical features that help estimate the relationship between SAV properties, cancer, and drug resistance. We developed a prediction system capable of distinguishing whether an SAV is associated with cancer, which achieved a five-fold cross-validation accuracy of 89.73%, a Matthews correlation coefficient of 0.74, and an F1 score of 0.81. Additionally, we constructed another machine learning model to predict mutations related to cancer drug resistance, which attained an accuracy of 86%, a Matthews correlation coefficient score of 0.57, and an F1 score of 0.66. Our innovative approach has great potential to serve as a valuable tool in cancer research and precision medicine.
Young-Jin Kim
Country : Korea
Official Title : 教授
Department :
Institute : Korea National Institute of Health
Speaker CVAdvancing Korean Population Genomics: Biobank-Scale Insights and Pangenome Initiatives
Recent advances in Korean population genomics have been propelled by large-scale biobank genotyping and pioneering pangenome projects. The Korea Biobank Array project analyzed genotype data from approximately 200,000 Korean samples and integrated these with international cohorts totaling nearly one million samples. This large-scale trans-ethnic meta-analysis uncovered hundreds of novel associations related to metabolic traits such as glycemic control, lipid profiles, liver enzymes, and serum urate. Polygenic risk scores derived from these findings effectively identified individuals at genetically high risk for diseases including type 2 diabetes and gout, with lifestyle factors further refining risk stratification. Parallel to these efforts, the Korean Pangenome Project is constructing high-quality telomere-to-telomere level genome assemblies from Korean volunteers using cutting-edge long-read sequencing technologies. Early drafts revealed extensive novel sequence variation not captured in current reference genomes, addressing the significant underrepresentation of East Asian populations in global pangenome resources. By 2029, the project aims to deliver 1,000 complete Korean pangenomes, creating an unparalleled resource for population-specific genomic research. Together, these initiatives lay a foundation for precision medicine tailored to Koreans and contribute critical insights into human genetic diversity and disease susceptibility.
Unveiling cellular and molecular mechanisms of brain developmental disorders using spatial transcriptomics
Brain developmental disorders such as lissencephaly and microcephaly arise from defects in fundamental processes including progenitor proliferation and neuronal migration. Despite rapid advances in genetics, the cellular mechanisms linking patient-derived variants to cortical malformations remain incompletely understood. In this talk, I will present our recent efforts applying spatial transcriptomics and integrative approaches to dissect disease mechanisms. We identified the first lissencephaly-associated variant in NDEL1 (p.R105P), which impaired nucleokinesis by disrupting its interaction with LIS1. Spatial transcriptomic profiling revealed complementary expression patterns of NDE1 and NDEL1 in progenitors and postmitotic neurons, highlighting their distinct yet cooperative functions in cortical development. In a separate study, we uncovered a BAIAP2 variant (p.R29W) in a patient with posterior-predominant lissencephaly. Spatial transcriptomic mapping demonstrated a posterior-high expression gradient of Baiap2 in the developing cortex, providing a mechanistic link between gene expression and regional disease severity. Functional studies confirmed migration and morphogenesis defects caused by this variant. Finally, I will briefly discuss our work on KIF23, a mitotic kinesin required for neural stem cell maintenance, underscoring the importance of mitotic regulation in cortical size control. Together, these studies demonstrate how spatial biology enables mechanistic insights into neurodevelopmental disorders and points toward precision medicine strategies for their treatment.
Whole Genome Sequencing for Rare Disease Diagnosis
Rare diseases are collectively common, affecting approximately 3% of the global population, yet their clinical heterogeneity and overlapping phenotypes often complicate timely and accurate diagnosis. Conventional genetic testing methods, such as targeted gene panels and whole-exome sequencing (WES), have significantly improved diagnostic yield but remain limited. These approaches primarily focus on coding regions and may miss pathogenic variants in regulatory regions, structural variants, repeat expansions, and mitochondrial genome alterations. Whole-genome sequencing (WGS) offers a comprehensive, unbiased assessment of the entire genome and has demonstrated superior performance in diagnosing genetically unresolved rare diseases. WGS enhances variant detection across diverse classes, including noncoding, structural, and complex rearrangements, thereby increasing diagnostic yield, particularly in patients with atypical presentations or negative WES results. Integration of WGS with advanced bioinformatics, transcriptomic and epigenomic data, and phenotype-driven interpretation further refines diagnostic accuracy and facilitates discovery of novel disease genes. Future trends include the implementation of WGS as a first-tier diagnostic test, incorporation of long-read sequencing for resolving repetitive or structurally complex regions, and the application of artificial intelligence to improve variant prioritization. As sequencing costs decline and data-sharing frameworks evolve, WGS is expected to become an essential tool for precision medicine, enabling earlier diagnoses, optimized management strategies, and novel therapeutic opportunities for patients with rare diseases.
top